Word-of-mouth only. No ads, no VC, no sales team. Just John, Atlas, and the network.
Word of mouth only works when someone sees something impressive and immediately wants to show it to the person next to them. Cooper's job is to be that thing — reliably, every time, for every audience.
The first screen a new user sees should be a fork: three doors that lead to completely different onboarding experiences. Same Cooper underneath, but the entry experience — the headline, the examples, the first prompt — speaks their language.
Student door: casual tone, examples about professors/internships/deadlines, visual energy that doesn't feel corporate. Work door: efficient, inbox-first, ROI framing ("never miss a follow-up"). Life door: warm, personal, memory-first ("your AI that never forgets"). Each persona gets a unique shareable link — Charlotte's friends share the Student link, John's colleagues share the Work link, the mom in the neighborhood gets the Life link.
When a user shares Cooper, they share the version that resonates with their world. Charlotte's college friends get the Student experience automatically. John's CRE peers get Work. The mom gets Life. Each door makes the referral multiplier go up because the first impression already speaks to that person's life — without John having to explain anything.
The grassfire only works if each user brings in 2+ others. Below 2 the fire goes out. Above 2 it compounds. Every product and process decision should be tested against one question: does this make a user more likely to show this to a friend?
John is at lunch. He pulls out his phone. The connect button doesn't work. Or it takes 12 seconds to respond. Or it says something generic. The buddy shrugs and that referral is dead forever — and John looks bad. One broken demo poisons 5 potential referrals. The Gmail connect flow must work perfectly on any phone, any time, with zero errors. This is Priority #1 above everything.
Kylie signs up. She tries it. It's decent. She closes the tab and never returns. No reason to come back — no notification, no proactive nudge, no "hey, you got 3 important emails this morning." Habit is what creates evangelists. One-time visitors create nothing. The agent needs to reach out first, not wait to be asked.
Charlotte's friends don't live in their inbox. An AI that summarizes email is exciting to a 50-year-old partner at a law firm. To a 20-year-old junior at UT, it's a shrug. The persona fork fixes this — but only if the Student experience is actually built around what students care about: professors, deadlines, internships, professional communication they're not used to doing yet.
John knows ~200–300 people well enough to personally recommend something. At 10% conversion, that's 20–30 users. If those users don't each bring in 2+ more, the fire goes out. The second-degree network must activate organically — which only happens if the product is impressive enough that users evangelize without being asked.
"Sign up as a favor" has near-zero referral multiplier. Charlotte's and John IV's friends need to experience something that solves a real problem they have — not just impress them for 2 minutes. The Student persona must be built around genuine student pain points to generate real evangelists, not polite signups.
John's peer network. Every conversation is a demo opportunity. But stop sending links — do live demos on the phone. Hand it to the person, let them type something. Set them up on the spot in 3 minutes if they're interested. Goal: 10 demos/week → 2–3 signups/week → 100 users in 10 weeks. Ceiling: ~200–300.
Charlotte and John IV each personally onboard 5 friends this week. Not via link — in person. Those 5 each have 200+ connections. One viral moment (a friend posts "my AI just emailed my professor and got me an extension lol") and you're at 1,000 from a single post. Requires the Student persona to be built first.
One vertical: CRE brokers and developers in Austin. John knows 50 personally. Each knows 50 more. Invite-only with referral codes — 3 invites per user. Scarcity creates status. Professionals spread differently: deal meetings, happy hours, industry events. Slower to 1,000 but higher quality users who are more likely to pay.
The grassfire ignites through demos, not links. Fix the demo first — then do 10 demos a week. The Three Doors (persona fork) is the unlock that lets each network spread naturally without John having to be in every conversation.